Homologous recombination deficiency (HRD) testing on cell-free tumor DNA from peritoneal fluid.

BACKGROUND: Knowing the homologous recombination deficiency (HRD) status in advanced epithelial ovarian cancer (EOC) is vital for patient management. HRD is determined by BRCA1/BRCA2 pathogenic variants or genomic instability. However, tumor DNA analysis is inconclusive in 15-19% of cases. Peritoneal fluid, available in > 95% of advanced EOC cases, could serve as an alternative source of cell-free tumor DNA (cftDNA) for HRD testing. Limited data show the feasibility of cancer panel gene testing on ascites cfDNA but no study, to date, has investigated HRD testing. METHODS: We collected ascites/peritoneal washings from 53 EOC patients (19 from retrospective cohort and 34 from prospective cohort) and performed a Cancer Gene Panel (CGP) using NGS for TP53/HR genes and shallow Whole Genome Sequencing (sWGS) for genomic instability on cfDNA. RESULTS: cfDNA was detectable in 49 out of 53 patients (92.5%), including those with limited peritoneal fluid. Median cfDNA was 3700 ng/ml, with a turnaround time of 21 days. TP53 pathogenic variants were detected in 86% (42/49) of patients, all with HGSOC. BRCA1 and BRCA2 pathogenic variants were found in 14% (7/49) and 10% (5/49) of cases, respectively. Peritoneal cftDNA showed high sensitivity (97%), specificity (83%), and concordance (95%) with tumor-based TP53 variant detection. NGS CGP on cftDNA identified BRCA2 pathogenic variants in one case where tumor-based testing failed. sWGS on cftDNA provided informative results even when tumor-based genomic instability testing failed. CONCLUSION: Profiling cftDNA from peritoneal fluid is feasible, providing a significant amount of tumor DNA. This fast and reliable approach enables HRD testing, including BRCA1/2 mutations and genomic instability assessment. HRD testing on cfDNA from peritoneal fluid should be offered to all primary laparoscopy patients.

Circulating Tumor DNA from Ascites as an alternative to tumor sampling for genomic profiling in ovarian cancer patients.

Genomic testing is crucial for the management of ovarian cancer. DNA from biopsies at diagnostic laparoscopies or interval debulking surgery after neoadjuvant chemotherapy, has a high failure rate. At relapse, biopsies may not be feasible. The aim of our study was to evaluate the feasibility and usefulness of measuring genomic instability score (GIS) on cell-free DNA (cfDNA) from ascites.Patients enrolled in a prospective study (NCT03010124) consented to analysis of biological samples. CfDNA was extracted from 1 to 4 ml of double-centrifuged fresh ascites. Targeted Next-generation sequencing (NGS) including TP53 mutation (TP53m) was performed on cfDNA to confirm the presence of tumor cfDNA. Single Nucleotide Polymorphism Array estimating somatic copy number alterations (SCNA) was performed to calculate GIS for Homologous-Recombination deficiency (HRD).Twenty nine ascites were collected from 20 patients with suspected or confirmed OC. 93% (27/29) samples had detectable cfDNA (median 1120 ng [24-5732]) even when obtained during chemotherapy. A deleterious mutation was identified in 100%, with high allelic frequencies (median 60% [3.3-87%]), confirming that cfDNA was tumoral. SCNA analyses on 17 patients showed 11 high GIS, and 6 low GIS. 4 patients with confirmed BRCA mutation had a high GIS on ascites. When available from the same patient, SCNA profiles on ascites and tumor were superimposable.Ascites is frequent at diagnosis and relapse and yields large amounts of tumoral cfDNA. SCNA analysis on ascitic cfDNA is feasible and can detect the same HRD scar as tumor testing. Ascites could provide an alternative to tumor sampling for HRD and BRCA testing.

Clinical Relevance of BRCA1 Promoter Methylation Testing in Patients with Ovarian Cancer.

PURPOSE: Homologous recombination deficiency (HRD) is closely related to PARP inhibitor (PARPi) benefit in ovarian cancer. The capacity of BRCA1 promoter methylation to predict prognosis and HRD status remains unclear. We aimed to correlate BRCA1 promoter methylation levels in patients with high-grade ovarian cancer to HRD status and clinical behavior to assess its clinical relevance. EXPERIMENTAL DESIGN: This is a retrospective monocentric analysis of patients centrally tested for genomic instability score (GIS) by MyChoice CDx (Myriad Genetics). The detection of BRCA1 promoter methylation and quantification of methylation levels were performed by quantitative droplet digital PCR methodology. High BRCA1 methylation was defined as ≥70% and deemed to be associated with homozygous silencing. RESULTS: Of 100 patients, 11% harbored a deleterious BRCA1/2 mutation. GIS was considered positive (score ≥ 42) for 52 patients and negative for 48 patients. Using a 70% cutoff, 19% (15/79) of BRCA wild-type ovarian cancer had high BRCA1 methylation levels. All of the highly methylated tumors were classified as HRD, achieving a positive predictive value of 100%. We detected 14% (11/79) low-methylated tumors (1%-69%), and all of them were also classified as HRD. Mean GIS was 61.5 for BRCAmut, 66.4 for high-BRCAmeth, 58.9 for low-BRCAmeth, and 33.3 for BRCAwt unmethylated (P < 0.001). Low methylation levels detected in samples previously exposed to chemotherapy appeared to be associated with poor outcome post-platinum. CONCLUSIONS: Patients with ovarian cancer with high levels of BRCA1 hypermethylation are very likely to have high GIS and therefore represent good candidates for PARPi treatment. These results may be highly relevant to other tumor types for HRD prediction. See related commentary by Garg and Oza, p. 2957.

The Identification of Large Rearrangements Involving Intron 2 of the CDH1 Gene in BRCA1/2 Negative and Breast Cancer Susceptibility.

E-cadherin, a CDH1 gene product, is a calcium-dependent cell-cell adhesion molecule playing a critical role in the establishment of epithelial architecture, maintenance of cell polarity, and differentiation. Germline pathogenic variants in the CDH1 gene are associated with hereditary diffuse gastric cancer (HDGC), and large rearrangements in the CDH1 gene are now being reported as well. Because CDH1 pathogenic variants could be associated with breast cancer (BC) susceptibility, CDH1 rearrangements could also impact it. The aim of our study is to identify rearrangements in the CDH1 gene in 148 BC cases with no BRCA1 and BRCA2 pathogenic variants. To do so, a zoom-in CGH array, covering the exonic, intronic, and flanking regions of the CDH1 gene, was used to screen our cohort. Intron 2 of the CDH1 gene was specifically targeted because it is largely reported to include several regulatory regions. As results, we detected one large rearrangement causing a premature stop in exon 3 of the CDH1 gene in a proband with a bilateral lobular breast carcinoma and a gastric carcinoma (GC). Two large rearrangements in the intron 2, a deletion and a duplication, were also reported only with BC cases without any familial history of GC. No germline rearrangements in the CDH1 coding region were detected in those families without GC and with a broad range of BC susceptibility. This study confirms the diversity of large rearrangements in the CDH1 gene. The rearrangements identified in intron 2 highlight the putative role of this intron in CDH1 regulation and alternative transcripts. Recurrent duplication copy number variations (CNV) are found in this region, and the deletion encompasses an alternative CDH1 transcript. Screening for large rearrangements in the CDH1 gene could be important for genetic testing of BC.